Steel Quotes in Seconds - The AI Playbook Every Metals Portco Can Copy-Paste

Six months, one AI strike-team, and a cloud-native tech stack turned a 50+ unit metal fabricator into an auto-quoting machine. If you run businesses where speed wins orders and RFQ delays erode margin, this playbook repeats.

In metals, the first price often wins. Yet this global fabricator's reps needed multiple hours to price a single request. Every 4.5 minutes a new quote landed by phone, email, or PDF, but legacy ERP screens couldn't surface inventory, mill cost, or ship dates fast enough. Prospects shopped elsewhere, sales churned on low-value admin, and IT, which lacked AI talent, while leadership feared an automation misfire more than lost deals.

We dropped in a two-person pod: one senior AI engineer, one product owner. Week 1 we mapped data flows and piped three years of orders, inventory, and freight tables into Azure. Weeks 2-4 produced a proof-of-concept that read unstructured emails with Azure OpenAI, pulled live stock and pricing via Logic Apps, and composed draft quotes in Outlook. By Week 6 the MVP was containerised on Kubernetes, queued by Service Bus, and live in production in North America.

Quotes that once took hours now arrive in minutes. Half require < 5 minutes of human polish before sending. That velocity alone is forecast to add $75M in marginal revenue during year one by converting prospects faster than competition. Reps now spend time negotiating, not data-entry, rookie sellers ramp faster, and customers praise the "Amazon-like" turnaround.

Becoming quoter of choice is a wedge. The same AI backbone can price cut-to-length plate, schedule mill slots, or auto-generate manufacturing certificates. Any PE portfolio company that lives or dies on response time can splice its data into this architecture and capture the same upside.

Want the playbook? Would love to chat.